Search results

Add Habana/Intel Gaudi support

enhancementNew feature or requestNew feature or request

Status: Open.

#1864 In kvcache-ai/ktransformers;

· albertschmidt092-ux opened on Feb 23, 2026

epyc 7402 CPU上推理Qwen3-235B-A22B出现illegal instruction

Status: Open.

#1863 In kvcache-ai/ktransformers;

· 108851027 opened on Feb 23, 2026

BF16 backend cannot parse gpt-oss packed/fused expert weight format

Status: Open.

#1861 In kvcache-ai/ktransformers;

· vshortt73 opened on Feb 22, 2026

# MXFP4 quantization allocates GPU weights for all experts regardless of KTransformers <code>--kt-num-gpu-experts</code>

Status: Open.

#1860 In kvcache-ai/ktransformers;

· vshortt73 opened on Feb 22, 2026

SGLang main passes <code>num_gpu_experts</code> but kt-kernel expects <code>gpu_experts_mask</code>

Status: Open.

#1859 In kvcache-ai/ktransformers;

· vshortt73 opened on Feb 22, 2026

KTransformers TaskQueue coordinator thread spins at 100% CPU when idle

Status: Open.

#1858 In kvcache-ai/ktransformers;

· Halcyonhal9 opened on Feb 21, 2026

请问一下现在KTRANSFORMERS支持的GGUF型号

Status: Open.

#1856 In kvcache-ai/ktransformers;

· tech112233445566 opened on Feb 21, 2026

启用DeepSeek-V3.2 GGUF sglang服务时，capture cuda graph失败

Status: Open.

#1854 In kvcache-ai/ktransformers;

· Zhai-Huichen opened on Feb 18, 2026

是否有计划支持视觉输入？如Qwen3.5?

enhancementNew feature or requestNew feature or request

Status: Open.

#1852 In kvcache-ai/ktransformers;

· FanDefa opened on Feb 17, 2026

FP8-origin Qwen3 models Serving Success but Inferring Failed

Status: Open.

#1847 In kvcache-ai/ktransformers;

· SchenbergZY opened on Feb 12, 2026

AscendNPU run Qwen3-235b error

Status: Open.

#1843 In kvcache-ai/ktransformers;

· zxt-triumph opened on Feb 11, 2026

MiniMax2.1输出很慢很慢很慢，结果也是乱码，麻烦指点一下，谢谢先

Status: Open.

#1842 In kvcache-ai/ktransformers;

· zcc2xj opened on Feb 11, 2026